Additive Manufacturing Data Fusion for Layer Quality Validation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Additive manufacturing processes, such as selective laser melting, face challenges in process reproducibility and predictability due to complex correlations between process parameters and structural properties, making it difficult to visualize and understand the reasons for process failures and quality issues in the manufactured components.
Innovation Solution
A method is developed to collect and evaluate data during the additive manufacturing process, modifying it to delete data representing insufficient structural quality and superimposing additional sensor data to support validation, enabling graphical representation of complex sensor data in three dimensions for deep process understanding and real-time monitoring, which creates a digital twin of the component for improved reproducibility and control.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If extensive sensor data are collected during additive manufacturing, then process understanding and quality control are improved, but data complexity and difficulty of visualization increase
Solution Approach 1:
The patent segments the complex sensor data into multiple hierarchical levels: process data (laser parameters, powder bed temperature), sensor data (acoustic emission, thermography, X-ray), and result data (structural properties). This segmentation allows systematic analysis and visualization of individual data types while maintaining their interrelationships, resolving the contradiction between comprehensive data collection and manageable complexity
Solution Approach 2:
The patent introduces temporal and spatial dimensions to data visualization by creating time-resolved process monitoring and spatially resolved structural analysis. Multi-dimensional data representations including 3D visualizations and temporal evolution plots transform complex multi-parameter data into intuitive visual forms, enabling comprehensive process understanding without overwhelming complexity
2Measurement precision
If manual data comparison is performed to understand process failures, then analysis accuracy is improved, but time consumption increases significantly
Solution Approach 1:
The patent implements automated feedback mechanisms that continuously monitor process parameters and sensor data, comparing them against reference values and previous successful builds. Statistical process control methods and machine learning algorithms automatically identify deviations and predict failures, providing real-time feedback that eliminates time-consuming manual analysis while maintaining high accuracy through systematic data comparison
Solution Approach 2:
The patent creates digital twins and virtual replicas of the additive manufacturing process by collecting and storing comprehensive process and sensor data. These digital copies enable virtual analysis and prediction of process outcomes without requiring physical manual inspection, significantly reducing analysis time while maintaining accuracy through repeated virtual comparisons against reference models
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach facilitates in-line process monitoring and control, allowing for reliable prediction of structural and mechanical properties, reducing the need for expensive non-destructive testing and preventing process deviations by providing a comprehensive understanding of the manufacturing process, enhancing the reproducibility and quality of additive manufactured components.
Implementation Method 1
melting, e.g. by the energy of a laser beam
Implementation Method 2
selective laser melting (SLM)
Implementation Method 3
capturing at least two layer images of the part during the manufacture via a detection unit which is configured to spatially resolved capture a measure which characterizes an energy input into the part
Data Source
AI summary
A method of providing a dataset for additive manufacturing includes collecting a first type of data for the dataset during the additive buildup of a at least one layer of a component to be manufactured, evaluating of the structural quality of the layer by the first type of data, modifying the first type of data in that fractions of the data representing an insufficient structural quality of the layer are deleted from the first type of data, and superimposing second type of data, to the first type of data, wherein the second type of data is suitable to support a validation of the structural quality of the as-manufactured component.